On March 31, a bipartisan commission led by former Senator Charles Robb (D-Va.) and federal appellate court Judge Laurence Silberman, a Republican, reported to President George W. Bush on what went wrong in the intelligence community when it failed to accurately assess that Iraq did not possess stockpiles of weapons of mass destruction (WMD). The White House-appointed commission also offered recommendations on improving overall U.S. intelligence performance.
For some, the report’s conclusions were bound to be of questionable value because the commission’s mandate ignored policymakers’ actions on the intelligence and the interaction between policymakers and intelligence. True solutions to intelligence performance, in this view, must be considered in a more strategic and holistic way, considering both the supply and the demand side of the ledger. In normal circumstances, intelligence is a supporting function, contributing analysis and occasionally unique and secret data to a policy deliberation that includes many other information inputs. It is quite rare that intelligence shoulders the burden of making a war or peace judgment, particularly when there is no evidence of an intention to launch a direct attack on the United States or its forces.
Still, it is sad but true that there is room for blame at more than one address; we need to study and come to terms both with an intelligence failure and a policy failure. Although the Robb-Silberman report deals only with the first topic, it does it well. Of all the reports breathlessly assessing intelligence failures and proposing to fix the problem, this one is the best in terms of understanding the intelligence profession and in terms of setting a tone of realism and even humility regarding what can be credibly promoted as solutions to a very complex set of problems. The report provides some unusual insight into the art and science of intelligence analysis, and its recommendations, although often not original or dramatic, make common sense. If implemented fully, it would make for a better intelligence process and product.
Equally compelling is the report’s understanding of how much of the failures and underperformance are caused at least in part by the way large bureaucracies behave. The commission benefited from having two university presidents as members, who reportedly were deeply interested in issues of organizational behavior. The report is more satisfying than some for acknowledging that very large, complex organizations inevitably create rules and checks and balances that over time impede the organization’s ability to achieve its core mission and objectives. This is surely true of the intelligence community, which seems to thrive on making processes and procedures more complex.
That is why it is troubling that this commission as well as the ones that preceded it say almost nothing about the size and complexity of the big intelligence machine the U.S. government has constructed over the decades. Each report pays homage to concepts such as “streamlining,” eliminating “stovepipes,” creating greater efficiencies, etc., but none says that a smaller community would almost certainly be a more successful one. Each time we add a new office or agency, do we dis-establish an old one? Of course not, say the “iron laws of bureaucratic behavior.” If the report had followed its own advice to “integrate and innovate,” it would have considered dis-establishing the CIA, since many of its original roles and missions will be transferred to the new director of national intelligence, and calling for a downsizing of the overall community, in the interest of achieving more of a real “community” of common interests and goals.
Intelligence and the Iraq Target
The core purpose of the commission was to look back at the Iraq case and figure out what went wrong and then to look forward at intelligence solutions. The Iraq case spans 200 pages of the 600-page report and is compared briefly to Libya, al Qaeda in Afghanistan, terrorism, and a section on Iran and North Korea, the judgments on which did not make it through the security screen to appear in the unclassified report.
The judgment on Iraq is stark and sobering: “We conclude that the Intelligence Community was dead wrong in almost all of its pre-war judgments about Iraq’s weapons of mass destruction. This was a major intelligence failure.” After this show-stopper, the report goes on to say with some insight and even empathy that the pre-war hypothesis that Saddam Hussein had such weapons was reasonable given his past behavior, but should not have been turned into a presumption. It says that it would not have expected the community to get it all right but, rather, less wrong. This is a subtle but important understanding of the limits of intelligence that many in the U.S. media and therefore the American public do not fully grasp.
In a recent public presentation, Silberman said it was a “grave, grave mistake” to go from a judgment of past behavior to a “90 percent certainty that he had weapons of mass destruction.” The report faults the community for not remembering or implementing its own tradecraft on the question of weapons of mass destruction. It failed to be sufficiently aggressive in questioning the bona fides of human sources; it became lax in questioning assumptions, red-teaming (the use of a parallel, independent analytic group to use different assumptions and presumably come to different conclusions), and considering alternative hypotheses; and it conducted a National Intelligence Estimate (NIE) process that appeared to treat some dissenting views as trivial and failed to vet some technical disputes thoroughly through available auxiliary analytic processes.
Techniques to invite first-rate scientists in universities and private laboratories to critique government analysis exist and are often used on issues less vital than the war and peace context in which analysis on Iraq was conducted, so it is particularly disturbing that the community seemed to ignore or neglect some of its own internal checks. Examples of the weak tradecraft apply to each of the WMD subcategories: the aluminum tubes were allegedly linked to the nuclear program, the mobile laboratory trucks to biological weapons, and the water trucks to chemical weapons. Several of these issues were in dispute at the time of Secretary of State Colin Powell’s testimony to the UN Security Council in February 2003.
The commission also acknowledges that, in terms of broad political analysis, assuming the worst from Hussein’s regime was a reasonable position to take but faults the analysts for taking that hypothesis and making it a premise that was no longer subject to scrutiny—again, a sign of poor tradecraft. At all levels of the analytic cadres, there was a comfortable consensus among analysts that Hussein’s regime was not capable of reform and was relentlessly ambitious to accumulate additional attributes of national power. This was based on more than a decade of experience watching a closed and cruel regime, with proven aggressive behavior toward Iran, Kuwait, and its own citizenry.
Long before the Bush administration came into office, analysts were on a kind of automatic pilot with respect to the fundamental behavior and attitudes of the regime in Baghdad. Workshops and exercises tried to anticipate new and different actions by the regime, but very few posited that Iraq was largely passive and hunkered down. It would have been startlingly counterintuitive for analysts to argue that the regime had been pacified; there was too much data on Iraqi noncompliance and defiance of international efforts to make such an alternative hypothesis credible. Surely, independent minded analysts occasionally voiced such opinions, but it would have taken some definitive new evidence to allow the subtle and collective group think to shift to a new bottom-line judgment. The commission report makes some straightforward, common sense recommendations on how to reinvigorate some of the checks and balances in the analytic process but sensibly judges that such alternative analysis mechanisms are an incomplete solution.
The commission’s treatment of the Iraq case has two shortcomings. First, the role of the UN Special Commission on Iraq (UNSCOM) as a source of vital insight into Iraqi activities is understated and underacknowledged in the report. In the early 1990s, UNSCOM relied heavily on leads provided by UN member states, the United States almost certainly the most active among them, in helping UNSCOM launch its work and identify the suspected weapons sites in Iraq. Over time, UNSCOM became a mature and competent organization, with accumulated knowledge of Iraq that surpassed that of any member state, including states that provided professional cadres as inspectors. The U.S. intelligence community has difficulty admitting how dependent it was on UNSCOM reporting and on the experience brought back by inspectors. In the mid- to late 1990s, UNSCOM began to articulate its doubts about the likelihood of remaining stockpiles of proscribed weapons; despite its own acquired distrust of Iraqi declarations, its own methodology was leading it to certain conclusions. Yet, within the intelligence community there was subtle instinct to be more skeptical than UNSCOM, to assume that a UN organization could not be as tough-minded as the United States. The absence of new collection meant the analysts had little basis on which to challenge UNSCOM or draw a different conclusion than UNSCOM’s experts.
This is a fascinating tale of the evolution of a UN body that actually became more competent than the capabilities of the United States. The power of knowledge shifted from the United States to UNSCOM, but it was difficult for the U.S. side of the equation to admit it. The commission report skims lightly over the changing dynamic between UNSCOM and U.S. intelligence and therefore understates the value that future UN inspections and perhaps a permanent inspectorate could have on the ability of the United States and the international community to stay smart in a changing proliferation environment. It is a topic worthy of more attention, a book perhaps that only Charles Duelfer can write.
Second, on the question of whether analysts were subjected to political pressure, the report insists that it left no stone unturned to get to the bottom of the allegations. It provided hotlines and means by which analysts could report anonymously, and it concluded quite forcefully that it found “no evidence of politicization of the…assessments concerning Iraq’s reported WMD programs…and no evidence of politicization even under the broader definition used by the CIA’s Ombudsman for Politicization,” which includes any “unprofessional manipulation of information and judgments…to please what those officers perceive to be policymakers’ preferences.”
The commission concedes that “there is no doubt that analysts operated in an environment shaped by intense policymaker interest,” but this formulation seems to understate the problem. Analysts are rewarded for working well with customers and for focusing on policy-relevant work. The administration was highly confident in its own analysis, and government analysts were frequently challenged to demonstrate that they could hold their own or make useful contributions to the strong personalities in the president’s team, personalities who did not hesitate to express their disagreement if not disdain for views that did not conform to their own. So, it would seem a serious shortcoming on the commission’s part not to think beyond the narrow question, “Did you ever change any language because of pressure,” to imagine the highly charged and stressful environment in which the community’s Iraq experts were working. This is not to absolve analysts from responsibility, but to question whether the commission’s own fine work will be marked by this perceived politically correct conclusion. The commission could have been bolder and more strategic in its understanding of this critical issue, even within the constraints set by its mandate.
The WMD Proliferation Challenge
The commission was also tasked with thinking beyond the Iraq case, and it attempted to derive lessons from the Iraq experience for continuing intelligence coverage of proliferation. It usefully points out that the proliferation problem is getting more difficult, given changes in technology, information networks, and the emergence of black-market nuclear networks such as those led by Pakistani scientist Abdul Qadeer Khan. At the same time, the threshold for intelligence performance is also rising. That is to say, if you thought Iraq was hard, just wait and see how much more difficult Iran and the next cases will be. The report judges that the intelligence community’s efforts have not kept pace with proliferation and will need to be more aggressive and innovative in the future. Many of the recommendations, however, are ideas that have long been accepted and at least partly implemented by the community, including very active outreach to nongovernmental experts in academia and in industry. The commission may have somewhat unrealistic expectations of what nongovernmental experts bring to the table in such exchanges. At best, a creative synergy occurs, and new insight is gained, but the selection of outside experts for such exchanges can be fraught with political correctness and can lead to different but also erroneous judgments.
The commission considered systematically all aspects of the intelligence business, from collection to analysis to information sharing and the special problems of integrating new parts of the federal system since the September 11 attacks. It faced the awkward situation of having to adapt its analysis midstream to the Intelligence Reform and Terrorism Prevention Act of 2004, which, when the commission was founded, faced poor prospects for becoming law. For political more than merit reasons, the bill passed and the president signed it into law, creating the new director of national intelligence position that will now be held responsible for the increasingly large and complex intelligence system. One infers from the report that many of its recommendations could have been implemented by the leaders of the old system, that the authorities of the director of central intelligence could have been strengthened, and that adding a new system on top of the old was not necessary or desirable.
Still, the commission took the new legislation into account. It noted that, under the law, the director of national intelligence has the authority to create a national nonproliferation center comparable to the National Counterterrorism Center (NCTC). The NCTC combines analysis and operational planning and is mandated by last year’s law. Interestingly, the commission recommended a different model for the proliferation problem. It believes the National Counter Proliferation Center should provide all-source intelligence to a policy-based “Counterproliferation Joint Interagency Task Force,” which would conduct interdiction activities and coordinate interagency and partner nations’ counterproliferation activities. The task force would therefore be broader than an intelligence organization, drawing on military and law enforcement capabilities and playing a supporting role to the U.S. Strategic Command, which in January 2005 was designated the lead military command for WMD issues.
This focus on interdiction and on the military role in a robust counterproliferation strategy lends itself well to nuclear delivery systems and some components for developing reprocessing and enrichment capabilities, but seems not a perfect match for the biological weapons challenge, which the commission determines is the hardest challenge of all. The report recommends more coordination with the biological sciences community. It is a worthy thought, but given new post-September 11, post-anthrax restrictions on biological research, it will take wise leadership and courage both in the private biological science community and in government to manage all the security and professional disincentives to such coordination.
The commission was also wise to take on one of the most dysfunctional parts of the way the intelligence agencies create barriers to sharing of key information, which often makes a mockery of the concept of “community.” The report challenges the concept of ORCON, or originator controlled, which permits an agency that has generated raw data or intelligence to determine its distribution. The commission stated clearly that a change in mindset is called for; the information is “owned” by the U.S. government, not a single intelligence agency. A more standardized and streamlined classification system, a more integrated process for providing clearances to intelligence community personnel, and a shift away from information “sharing” to information “integration” would be important improvements in culture and procedures.
Less persuasive was the commission’s recommendation to create a new cadre of open-source information experts, even an “Open Source Directorate.” Again seeming to neglect their own mantra of innovation and integration, a better solution would be to train everyone in the analytic structures in better utilization of all-source information. To create a compartmented approach to the information that most concede probably constitutes well more than 90 percent of the total information needed for finished intelligence production seems to undermine their own beliefs.
Other enduring myths and practices in the intelligence community cannot easily be overcome. It is difficult to fix a culture that was created to warn the president and prevent future Pearl Harbors. Given the self-conscious promotion by intelligence community leaders of an ethos that distinguishes intelligence from other national security disciplines, many analysts begin to default to a warning function in their analysis. They are rarely discouraged from speculating on worst-case scenarios because they see their core mission as helping policymakers prepare for dangers and threats. It is counterintuitive to most trained analysts to speculate that things might be better than they appear. In the Iraq case, for example, it is easy to imagine that analysts, sensing that war was the likely choice of the president, gave their best shot at describing the risk environment in which U.S. forces would be sent, firm in the belief that warning was part of their unique role.
The NIE process, which has now been so fully scrutinized by the press and every commission, is never perfect but is also difficult to replace with something reliably better. Estimates vary in their utility to policymakers, not only because of how they are crafted, but because of different ways policy customers use them—back again to the structural limits of the commission’s mandate. Estimates that take into account all of the alternative analytic processes and that benefit from all possible collection initiatives will risk not being timely for decisions on fast-changing issues or being too complex in offering multiple outcomes to be easily absorbed by policymakers.
Fixing the Debate About Intelligence
Beyond the daunting task of reforming an excessively complex bureaucratic system, fixing intelligence also means fixing the debate about intelligence, i.e., getting public expectations into more realistic boundaries. In our increasingly information-saturated open society, everyone is entitled to an opinion or two about what’s wrong with government and why the fools in Washington keep doing such dumb things. Over the past few years, the sequence of intelligence failures well documented in the press and increasingly in academic literature creates the impression of a well-informed debate over intelligence issues. This is desirable, and surely many thoughtful citizens are better informed and can take responsible positions on intelligence reform and can convey such preferences to their elected officials.
Yet, there is also a cost to the frenzy of public attention, often in the form of ridicule, toward intelligence issues. The public debate has benefited from former officials becoming more comfortable talking about their careers in intelligence, and memoirs and even works of fiction by former intelligence officers appear more and more frequently. The ease with which matters once considered sensitive and secret can be discussed openly with no fear of penalty, however, means that policymakers should be less confident that intelligence matters will be kept in that special channel.
This can have—indeed may already have had—a chilling effect on whether policymakers request intelligence reports such as NIEs on the most sensitive topics or avoid them for fear of leaks. The erosion of public confidence in intelligence performance has also almost certainly made it more difficult for intelligence leaders to present information about new emerging threats to policymakers without careful scrutiny and questioning of the information. That in and of itself may be a healthy thing and gets us to the place where policymakers and intelligence share responsibility for making judgments that lead to policy decisions, but it could also lead to delays or reluctance to act for fear of unintended consequences.
The commission worked hard to untangle many of the complexities of modern intelligence but could not resolve the most difficult part of all, which is how publicly accountable policymakers use intelligence, particularly in circumstances where war is a choice but not the only option.
Ellen Laipson is president and chief executive officer of the Henry L. Stimson Center. She held various foreign policy and national security positions in a 25-year U.S. government career, including serving as vice chairman of the National Intelligence Council from 1997-2002.
1. Report of the Commission on the Intelligence Capabilities of the United States Regarding Weapons of Mass Destruction, March 31, 2005, p. 6, available at http://www.wmd.gov/report/wmd_report.pdf (hereinafter commission report).
6. Greg Treverton points out that, given the limits of intelligence, a doctrine of pre-emption to eliminate an adversary’s weapons of mass destruction essentially means eliminating the adversary. Gregory F. Treverton, “Intelligence: The Achilles Heel of the Bush Doctrine,” Arms Control Today, July/August 2003, pp. 9-11.